deriv3 modification to use gamma()

Rich Townsend (townsend@ruddy.cqs.washington.edu)
Thu, 22 Jan 1998 16:08:04 -0800


Here's a problem that I've been banging my head against the wall on:

I've got a nasty likelihood involving factorials. Since the gamma function
is equal to (x-1)!, I thought I would be able to modify the D and make.call
functions as in Venables & Ripley (2nd Ed.), on page 288, to handle the gamma
function, minimize my -log likelihood and calculate the variances on the
estimates. Unfortunately, I think what I need to do is modify the "dnorm"
and "pnorm" extensions of D and make.call with "gamma" and a derivative
function of gamma. Anyone have a good solution for the derivative handy?

Rich

Columbia Basin Research, Puget Sound Plaza, suite 1820
Seattle, Washington 98101 (206)616-7492
http://www.cqs.washington.edu/~townsend/jwz/index.shtml
-----------------------------------------------------
Never take life seriously. Nobody gets out alive anyway.